How to Design a GeoDB GEOG 419/519: Advanced GIS
© Arthur J. Lembo, Jr. Cornell University
Key is asking the right questions: • How can GIS technology be implemented to streamline existing functions or change the way we achieve a goal • What data will benefit the organization the most? • What data can be stored? • Who is responsible for maintaining the database? © Arthur J. Lembo, Jr. Salisbury University
Truths about Design • Time-consuming • No end-use applications • If not done:
10000 9000 8000 7000 6000
Production
5000 4000 3000
Implementation
2000 1000 0
Concept
Concept
– Database may not meet requirements – Can end up with duplicate, missing, or unnecessary data – Lack of necessary management © Arthur J. Lembo, Jr. techniques Salisbury University
Analysis
Design
Design
Data Collection
Implementation
Objectives of Design • Design should define goals, identify, analyze, and evaluate design alternatives, and create implementation plan • Investment of time and money up front saves even more time and money later • A functional, well-organized database: – – – – –
Satisfies organizational objectives Contains all necessary data Accommodates different views of the data Distinguishes maintenance apps from user apps Organizes data so that different users access the same data © Arthur J. Lembo, Jr. Salisbury University
© Arthur J. Lembo, Jr. Salisbury University
Design Guidelines • • • • • • • • • •
Involve users Take it one step at a time Build a team Be creative Create deliverables Keep organizational goals and objectives in focus Do not add detail prematurely Document carefully Be flexible Plan from you model © Arthur J. Lembo, Jr. Salisbury University
•
Create from scratch: create schema for features datasets, classes, and attribute tables. • Import Existing Data: a schema is created, and existing data in imported in. • Use CASE tools: computer aided software engineering principles and GUIs can be used to create a geodatabase • THIS IS HOW TO BUILD A COMPUTER DATABASE, BUT DOESN’T MEAN ITS OF ANY USE – for this you need good database © ESRI, Building Geodatabases design © Arthur J. Lembo, Jr. Salisbury University
Steps in Building a GeoDB • • • • •
Model the user’s view of data Define objects and relationships Select geographic representation Match to geodatabase elements Organize geodatabase structure
© Arthur J. Lembo, Jr. Salisbury University
Designing GIS Databases • • • • •
Model the users’ view Define entities and their relationships Identify representation of entities Match to GIS data model Organize into geographic data sets The first three steps develop the conceptual model, classifying features based on an understanding of the data required to support the organization’s functions, and deciding their spatial representation. The last two steps develop the logical model, matching the conceptual models to ArcGIS geographic data sets © Arthur J. Lembo, Jr. Salisbury University
The Data Model • Data Model is a formal definition of the data required in a GIS. Types include: – Structured List – Entity Relationship Diagram
• Purpose of the data model is to ensure that the data is identified and described in a completely rigorous and unambiguous fashion
© Arthur J. Lembo, Jr. Salisbury University
Example Structured List © ESRI
Feature
Layer
Type
Prim. Attr.
Owner
Landuse
LU
Poly
LUID
Planning
Soils
Soil
Poly
SoilID
na
Elevation
DEM
Raster na
na
Hydrography
Hydro
Line
ID
DEC
Roads
CL
Line
ID
Eng.
Buildings
Bldg
Poly
ID
Eng.
Parcels
Parcel
Poly
SBL
Assessor
© Arthur J. Lembo, Jr. Salisbury University
Model the User’s View • Identify the functions that support the organization’s goals and objectives • Identify the data required to support the functions • Organize the data into logical sets of features • Define an initial implementation plan • Identify organizational functions © Arthur J. Lembo, Jr. Salisbury University
© Arthur J. Lembo, Jr. Salisbury University
Define Entities and Relationships • Identify and describe entities • Identify and describe the relationships among these entities • Document the entities and relationship with UML diagrams • Create statements about how the system works then evaluate those statements in terms of entities and relationships © Arthur J. Lembo, Jr. Salisbury University
© Arthur J. Lembo, Jr. Salisbury University
The Nature of Geographic Data • Geographic data has been described as: – Object - a thing that can be seen or touched. – Entity - objects or things to be included in a database – Feature - the make, shape, form or appearance of a person or thing. Term that derives itself from cartography (features on a map) – Attribute - characteristics of the entities © Arthur J. Lembo, Jr. Salisbury University
Identify the representation of entities • Is the feature represented on a map? • Is the shape of a feature important? • Is the feature best accessed through its relationship with another feature? • Will the feature have different representations at different scales?
© Arthur J. Lembo, Jr. Salisbury University
© Arthur J. Lembo, Jr. Salisbury University
Classical Entities and Spatial Component
© ESRI
Entity Entity Attributes
Spatial Component
Street (name, number)
line (coordinates, topology)
Soil Zone (name, perm.)
polygon (coordinates, topology)
Well (type, date tested)
point (coordinates)
© Arthur J. Lembo, Jr. Salisbury University
E-R Diagram Platform
Side Number
Part of
Station © Arthur J. Lembo, Jr. Salisbury University
Name Zone
Spatial Relationships Spatial Relationship
Verbs
Connectivity
Connect, link
Contiguity
Adjacent
Containment
Contained
Proximity
Nearest
Coincidence
Coincident © Arthur J. Lembo, Jr. Salisbury University
Symbol
Considerations in Modeling Geography in an E-R Diagram • Correct Identification and Definition of Entities • Defining a Corresponding “Spatial” Entity for Each “Traditional” Entity • Recognition of Multiple Instances of Geographic Entities – time and scale – Represented by: entity simple, entity spatial, entity time © Arthur J. Lembo, Jr. Salisbury University
Representation of Spatial Objects Object Name
Object (entity) Spatial Object
Associated Spatial Object Type
G
T
XY Coordinate system © Arthur J. Lembo, Jr. Salisbury University
Topology: point, line, polygon
Modeling Spatial Relationships ENTITY
RELATIONSHIP
ATTRIBUTE © Arthur J. Lembo, Jr. Salisbury University
Developing a Spatial E-R Diagram • Is Derived From Needs Assessment • Relationships Determined from Application Descriptions Soil Sample Point
G
T
Contained In /Contains
© Arthur J. Lembo, Jr. Salisbury University
Farm Plot Polygon
G
T
Match to a GeoDB model • Determine appropriate geodb representation for entities – Spatial type = point • Unconnected – point feature • Connected – simple junction • Connected with internal topology – complex junction
– Spatial type = line • Stand-alone line = line feature • Line in a system = simple edge • Line with connected sections = complex edge
– Areas = polygon feature (with potential planar topology) – Objects = objects © Arthur J. Lembo, Jr. Salisbury University
© Arthur J. Lembo, Jr. Salisbury University
Physical Design • Logical Design Performed Independent of Physical Design • Must Now Move Logical Design to Physical Design • Physical Design Example (ARC/INFO, geodatabase):
© Arthur J. Lembo, Jr. Salisbury University
Conceptual vs. Physical Design Temp Logger G
Point
LINKS with T
Line
Physical Design Stream
PAT
BND
AAT
Stream Segment G
T
Stream
Reach Order
INFO
TIC
NAT
TempLog
ID
© Arthur J. Lembo, Jr. Salisbury University
History
ID Date Temp
Organize into Geographic Datasets • Assign entities to feature classes and subtypes – Feature classes vs. subtypes
• Group related sets of features into geometric networks or planar topologies – Simple edges & junctions, complex edges & junctions – geometric network – Need space-filling and no overlapping – planar topology
• Organize feature classes and datasets into geodb’s © Arthur J. Lembo, Jr. Salisbury University
© Arthur J. Lembo, Jr. Salisbury University